This paper describes the unexpected challenges of team tutor development such as the task and logistics. Previously, a research team from Iowa State University (ISU) working with the U.S. Army Research Laboratory (ARL) developed the reconnaissance (Recon) task for simple team tutoring with the Generalized Intelligent Framework for Tutoring (GIFT) (Bonner et al., 2015; Gilbert et al., 2015). Considerations were included for the testing environment such as audio-based team interactions, initialization of the scenario simultaneously, and the inclusion of eyetracking and screen capture technology. Throughout the process of tutor development, several computational challenges have been encountered such as the implementation of team rules, determination of the appropriate amount of feedback, and the use of participants’ behavior history as input to the tutor. Our descriptions of these challenges should forewarn future developers of team tutors. We also suggest enhancements to GIFT to aid this process.